330 Mr. H. F. Witherby on a Collect'ing [Ibis, 



Portugal. In the precediug table I have set out a com- 

 pari.sou o£ the geographical races o£ some of the resident 

 species in these two regions, the Oantabrian Mountains and 

 the Pyrenees. 



It will be seen that of these: (a) three are the same in the 

 Pyrenees, Cantabrians^ and north Portugal, and that all these 

 three are different in south Spain ; (J)) five, the same in the 

 Pyrenees and Cantabrians, are different in Portugal ; and (c) 

 six, the same in the Cantabrians and Portugal, are different 

 in the Pyrenees. 



Those under (Jj) and [c) especially form a puzzling mixture 

 of forms. Wliile I have no satisfactory explanation of these 

 ap})arent anomalies to offer, it may be of some interest to 

 point to the following considerations. The Oantabrian 

 Mountains are practically continuous with the Pyrenees and 

 spurs of the chain run down from its western end into north 

 Portugal. Thus there seems no definite isolatinii; barrier 

 between northern Portugal and the Pyrenees. Yet in group (h) 

 we have four races (Blue Tit, Crested Tit, Robin, and Stone- 

 chat) which are distinctly more richly coloured and darker 

 and two of them smaller in Portugal than in the Cantabrians 

 and Pyrenees, while another (Nuthatch) is less richly 

 coloured but also smaller. In group (c) we have four birds 

 (Jay, Goldfinch, Coal-Tit, and Dipper) in Portugal and the 

 Cantabrians which are distinctly darker (and in the Gold- 

 finch and Coal-Tit smaller) than those in the Pyrenees. 



In studying these differences all possible factors should be 

 considered, and as it has often been stated that humidity 

 plays a part, I subjoin a table of the values of the relative 

 humidity at various stations, which has been most kindly 

 supplied to me by the Meteorological Office, Air Ministry. 

 The figures in this table indicate the ratio of aqueous vapour 

 in a measured volume of air to the amount which the volume 

 would contain if the air were saturated. The figures for 

 each month have been supplied, but as the monthly variations 

 at the different stations are fairly similar I give only the 

 yearly mean. 



